Your RSA-2048 keys break in 2030. Find every one of them before attackers do.
🐍
🐍 PyPI
Not in CISA KEV
LOW severity

GHSA-m34j-p8rj-wjxq

LOWFix: tensorflow/tensorflow@6778470

GHSA-m34j-p8rj-wjxq is a low-severity (CVSS 2.5) CWE-369 vulnerability in tensorflow. 1 public exploit reference exists, so weaponization risk is real. O3 Security confirms whether GHSA-m34j-p8rj-wjxq is actually reachable in your code before you act, and blocks exploitation at runtime until you patch.

Division by 0 in `QuantizedBiasAdd`

Also known asBIT-tensorflow-2021-29546CVE-2021-29546PYSEC-2021-183PYSEC-2021-474PYSEC-2021-672
Published
May 21, 2021
Updated
Mar 13, 2026
Affected
12 pkgs
Patched
12 / 12
Exploits
1 known
Exploitation data as of Mar 13, 2026 · OSV.dev, NVD, FIRST.org (EPSS)

Real-World Exposure

12 pkgs affected
🐍tensorflow🐍tensorflow🐍tensorflow🐍tensorflow🐍tensorflow-cpu🐍tensorflow-cpu🐍tensorflow-cpu🐍tensorflow-cpu+4 more

Real-time download stats are indexed for npm and PyPI packages. This vulnerability affects PyPI packages — download data is not available via public APIs for these ecosystems.

Description

Impact

An attacker can trigger an integer division by zero undefined behavior in tf.raw_ops.QuantizedBiasAdd:

import tensorflow as tf

input_tensor = tf.constant([], shape=[0, 0, 0, 0], dtype=tf.quint8)
bias = tf.constant([], shape=[0], dtype=tf.quint8)
min_input = tf.constant(-10.0, dtype=tf.float32)
max_input = tf.constant(-10.0, dtype=tf.float32)
min_bias = tf.constant(-10.0, dtype=tf.float32)
max_bias = tf.constant(-10.0, dtype=tf.float32)

tf.raw_ops.QuantizedBiasAdd(input=input_tensor, bias=bias, min_input=min_input,
                            max_input=max_input, min_bias=min_bias,
                            max_bias=max_bias, out_type=tf.qint32)

This is because the implementation of the Eigen kernel does a division by the number of elements of the smaller input (based on shape) without checking that this is not zero:

template <typename T1, typename T2, typename T3>
void QuantizedAddUsingEigen(const Eigen::ThreadPoolDevice& device,
                            const Tensor& input, float input_min,
                            float input_max, const Tensor& smaller_input,
                            float smaller_input_min, float smaller_input_max,
                            Tensor* output, float* output_min,
                            float* output_max) {
  ...
  const int64 input_element_count = input.NumElements();
  const int64 smaller_input_element_count = smaller_input.NumElements();
  ...
  bcast[0] = input_element_count / smaller_input_element_count;
  ...
}

This integral division by 0 is undefined behavior.

Patches

We have patched the issue in GitHub commit 67784700869470d65d5f2ef20aeb5e97c31673cb.

The fix will be included in TensorFlow 2.5.0. We will also cherrypick this commit on TensorFlow 2.4.2, TensorFlow 2.3.3, TensorFlow 2.2.3 and TensorFlow 2.1.4, as these are also affected and still in supported range.

For more information

Please consult our security guide for more information regarding the security model and how to contact us with issues and questions.

Attribution

This vulnerability has been reported by Yakun Zhang and Ying Wang of Baidu X-Team.

Affected Packages

12 total 12 fixed
EcosystemPackageVulnerable rangeFix
🐍PyPItensorflowall versions2.1.4
🐍PyPItensorflow2.2.0&&< 2.2.32.2.3
🐍PyPItensorflow2.3.0&&< 2.3.32.3.3
🐍PyPItensorflow2.4.0&&< 2.4.22.4.2
🐍PyPItensorflow-cpuall versions2.1.4
🐍PyPItensorflow-cpu2.2.0&&< 2.2.32.2.3
Exploits & PoCs
1

Research use only. For defensive security, authorized penetration testing, and academic research only. Never execute exploit code against systems without explicit written authorization.

Detection & mitigation playbook

Open-source dependency
  1. Detect

    Scan your dependency tree (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, etc.) for tensorflow. O3's reachability analysis confirms whether the vulnerable code path is actually invoked in your application, so you act on real exposure instead of every transitive match.

  2. Fix

    Update tensorflow to 2.1.4 or later, then make sure no transitive (indirect) dependency still pins the vulnerable range — O3 confirms GHSA-m34j-p8rj-wjxq is resolved across your whole dependency graph.

  3. Workarounds

    If you can't upgrade right away: gate or disable the affected feature, validate untrusted input at the boundary, and avoid passing attacker-controlled data into the vulnerable path. O3's runtime protection blocks exploitation in production as an interim safeguard until the upgrade lands.

  4. How O3 protects you

    O3 pinpoints whether GHSA-m34j-p8rj-wjxq is reachable in your code and exactly where to fix it, then blocks exploitation in production at runtime until the patched version is deployed.

Tailored to GHSA-m34j-p8rj-wjxq. Runtime protection reduces exposure until a permanent patch is applied and verified — it complements patching, it doesn't replace it.

Frequently Asked Questions

### Impact An attacker can trigger an integer division by zero undefined behavior in `tf.raw_ops.QuantizedBiasAdd`: ```python import tensorflow as tf input_tensor = tf.constant([], shape=[0, 0, 0, 0], dtype=tf.quint8) bias = tf.constant([], shape=[0], dtype=tf.quint8) min_input = tf.constant(-10.0, dtype=tf.float32) max_input = tf.constant(-10.0, dtype=tf.float32) min_bias = tf.constant(-10.0, dtype=tf.float32) max_bias = tf.constant(-10.0, dtype=tf.float32) tf.raw_ops.QuantizedBiasAdd(input=input_tensor, bias=bias, min_input=min_input, max_input=max_input, min_b
O3 Security · Impact-Aware SCA

Is GHSA-m34j-p8rj-wjxq in your dependencies?

O3 detects GHSA-m34j-p8rj-wjxq across PyPI dependencies and uses function-level reachability to confirm whether the vulnerable code path is actually reachable — not just present. No false positives.

GHSA-m34j-p8rj-wjxq: Division by 0 in… | O3 Security